Linked Data Applications for Library Collections

28
Linked Data Applications for Library Collections Elk Oswood

Transcript of Linked Data Applications for Library Collections

Page 1: Linked Data Applications for Library Collections

Linked Data Applications for Library Collections

Elk Oswood

Page 2: Linked Data Applications for Library Collections

Linked Data and the Semantic WebSemantic Web (SW) Project

● Tim Berners-Lee (2006 W3C)

● Extend Web to make data machine-readable

Linked Data

● Set of best practices for publishing data on the Web.

Page 3: Linked Data Applications for Library Collections

Talk Outline1. Overview of Relevant Web Technologies

2. Overview of Linked Data

3. Applications of Linked Data for Library Use

4. Conclusion and Discussion

Page 4: Linked Data Applications for Library Collections

1. Overview of Relevant Web Technologies

Page 5: Linked Data Applications for Library Collections

1. Overview of Relevant Web TechnologiesRelevant Web Technology:

● Hypertext Transfer Protocol (HTTP)

● Uniform Resource Identifier (URI)

Page 6: Linked Data Applications for Library Collections

Hypertext Transfer Protocol (HTTP)● Request-Response

protocol

● Client sends requests

for resources.

● Servers send responses

back to client.

Page 7: Linked Data Applications for Library Collections

Uniform Resource Identifier (URI)● Unique sequence of characters

● Describes resource

● Uniform Resource Locator (URL) is a specific type of identifier

○ Tells how to access resource

○ HTTP

Page 8: Linked Data Applications for Library Collections

2. Overview of Linked Data

Page 9: Linked Data Applications for Library Collections

2. Overview of Linked Data● Set of Principles

○ Name starts with HTTP

○ Looking up an HTTP name returns useful

data.

○ Anything else with a relationship also has

name beginning with HTTP.

https://ontola.io/what-is-linked-data/

Page 10: Linked Data Applications for Library Collections

Linked Open Data● Open data concepts

● Allows data to be accessible to any human or machine with access to the Internet.

● No paywall, copyright, or patent rights.

Page 11: Linked Data Applications for Library Collections

Applications of Linked Data for Library Use

Page 12: Linked Data Applications for Library Collections

Linked Open Citation Database (LOC-DB)● Project created by researchers Anne Lauscher and colleagues.

● Current system:

○ Databases (DBs) of citations that are closed to other DBs.

● Goal: Create a Database of Citations

○ Designed for use in libraries

○ Citations taken from references within a work.

○ Linked Open Data Concepts

○ Designed to foster collaboration and sharing.

Page 13: Linked Data Applications for Library Collections

LOC-DB● Semi-automated

● Distributed System

○ Separate Instances of DB

● Resource Metadata Creation:

○ Union Catalogues

○ External Suggestions

● Automatic Reference Extractor (ARE)

● Editorial System

○ Ingestion

○ Reference Linking

○ Metadata Editing

Source 1

Page 14: Linked Data Applications for Library Collections

ARE Text-driven Reference Extraction

● Optical character recognition (OCR)

○ Open source software OCRopus

● ParsCit

○ Open source package

○ Detect and label citation strings from text.

Source 1

Page 15: Linked Data Applications for Library Collections

ARE Layout-driven Extraction

Source 1

Source 1

Page 16: Linked Data Applications for Library Collections

User Study and Quantitative Evaluation● Stakeholder Input:

○ Union Catalogues

○ Workflow issues

● Findings:

○ Scanning

○ Reference Linking

Source 1: Minimum, maximum, and median time in seconds

for the reference linking step with a sample size of 444.

Page 17: Linked Data Applications for Library Collections

Author’s Take-awaysPros:

● Open Data

● Both born digital and scanned documents

● Semi-automated component

○ Higher quality

○ Better results.

Cons:

● Current high cost to begin population of

LOC-DB.

● No conclusive group to host LOC-DB.

● Many libraries needed.

Page 18: Linked Data Applications for Library Collections

The Novel Authoritative Interlinking of Schema and Concepts (NAISC)● Researchers Lucy McKenna, Christophe Debruyne, Declan O’Sullivan

● Initial survey found:

○ Linked Data tools designed specifically for Information Professionals were desired.

● What they wanted:

○ Better tools designed specifically for libraries.

○ Work with the current workflow

Page 19: Linked Data Applications for Library Collections

NAISC Interlinking Framework

Source 3

Page 20: Linked Data Applications for Library Collections

Provenance● Information about

○ People

○ Institutions

○ Resources

○ Processes

Source 3: NAISC GUI: Partial Provenance GraphSource 3: NAISC Portion of GUI: Entity Interlinking

Page 21: Linked Data Applications for Library Collections

NAISC Interlinking Framework

Source 3Source 3

Page 22: Linked Data Applications for Library Collections

User Evaluation

Source 3: NAISC Think Aloud Evaluation

Page 23: Linked Data Applications for Library Collections

Author’s Take-AwaysPros:

● Interlink across datasets

● Facilitates increased engagement in Linked Data interlinking process.

Cons

● Current GUI

● Time consuming

Page 24: Linked Data Applications for Library Collections

Discussion

Page 25: Linked Data Applications for Library Collections

AcknowledgementsThank you to Elena Machkasova for her feedback and guidance.

Page 26: Linked Data Applications for Library Collections

Questions

Page 27: Linked Data Applications for Library Collections

References

Page 28: Linked Data Applications for Library Collections

References1. A. Lauscher, K. Eckert, L. Galke, A. Scherp, S. T. R.Rizvi, S. Ahmed, A. Dengel, P.

Zumstein, and A. Klein. Linked open citation database: Enabling libraries to

contribute to an open and interconnected citation graph. In Proceedings of the

18th ACM/IEEE on JointConference on Digital Libraries, JCDL ’18, page109–118,

2018.

2. L. McKenna, C. Debruyne, and D. O’Sullivan. Understanding the position of

information professionals with regards to linked data: A survey of

libraries,archives and museums. In Proceedings of the 18th ACM/IEEE on Joint

Conference on Digital Libraries,JCDL ’18, page 7–16,, 2018.

3. L. McKenna, C. Debruyne, and D. O’Sullivan. NAISC:An authoritative linked data

interlinking approach forthe library domain. In2019 ACM/IEEE JointConference

on Digital Libraries (JCDL), pages 11–20, 2019.